Professional Documents
Culture Documents
4 Ieee Asme
4 Ieee Asme
5, OCTOBER 2009
Abstract—Waveform design methods for piezo inkjet dispensers sure waves. However, an in-depth study of meniscus motion
based on measured meniscus motion are presented. The meniscus behavior might be difficult because the meniscus measurement
motion is measured from charge-coupled-device camera images in the paper by Bogy and Talke relied on postprocessing a
wherein strobe lights from light-emitting diodes are synchronized
with the jetting signal. Waveforms for the piezo dispenser are prohibitively large number of charge-coupled-device (CCD)
designed such that the number of experiments can be signifi- camera images.
cantly reduced compared to conventional methods. Furthermore, Recently, the meniscus motion in an inkjet nozzle was mea-
the designed waveform can also be evaluated by the measured sured using particle tracking velocimetry (PTV) or particle
meniscus motion since the motion is directly related to jetting image velocimetry (PIV) techniques [10], [11]. Using visual-
behavior. [2009-0032]
ization, the jetting behavior can be understood. However, addi-
Index Terms—Measurement. tional particles, which might change jetting performance, must
be included for PIV measurement. Also, special equipment
I. I NTRODUCTION is required for particle velocity measurement. In summary,
measurement methods based on PIV or PTV may not be quick
I NKJET technology has recently emerged as one of the
most powerful tools for patterning electronic applications
such as large-area displays, radio-frequency identification, and
and simple enough for meniscus motion measurement and
waveform design in industrial printing devices.
printed-circuit-board patterning [1]–[3]. As inkjet applications In this paper, a new method for measuring meniscus motion
broaden, various types of jetting materials are required to be is proposed. The proposed technique uses a CCD camera sys-
precisely dispensed from the inkjet head. Jetting fluids need to tem wherein light-emitting-diode lights are synchronized with
have jettable material properties, as discussed in [4] and [5]. the jetting signal. Therefore, the proposed method does not
In addition, jetting performance needs to be well controlled require special hardware, unlike previous methods using PIV
for inkjet technology to be viable in various applications. To techniques. Most printing systems already have a CCD camera
control jetting performance parameters such as droplet speed system with a strobe LED for measuring droplet speed [1],
and droplet volume, the input waveform voltage driving the [7]–[9]. Therefore, only software algorithms need to be imple-
printhead must be properly designed [6]–[9]. In practice, a mented in existing CCD camera systems for in situ measure-
simple trapezoidal waveform has been widely used for this ment of meniscus motion. The proposed software algorithm has
purpose. To find the optimal dwell time for the waveform, the two parts: 1) image processing to identify the meniscus location
relationship between droplet jetting speed and dwell time has from the CCD camera image and 2) controlling LED lights
been sought [6]–[8]. However, this conventional approach for that enable CCD camera image acquisitions of the meniscus
designing the waveform requires a number of experiments to motion over the duration of interest after the jetting signal. The
scan the dwell time for its relationship to droplet jetting speed. measurement of meniscus motion has the advantage that jetting
Furthermore, if there is no jetting from inkjet dispensers, it is behavior can be predicted without actual jetting.
difficult to understand the possible causes of nonjetting. In such As an application of the measured meniscus motion, a wave-
cases, a trial-and-error approach has been the only option to form design method is presented. It should be noted that the
determine the jetting conditions. In this paper, a new waveform motion is likely to be complex because it is the combined
design method is proposed such that time-consuming efforts to result of pressure waves generated from both the rising and
find optimal waveforms can be reduced. falling sections of the waveform. For straightforward analysis
Since the experimental study by Bogy and Talke [7], jetting of meniscus motion behavior in relation to a waveform, the use
phenomena have been known to be related to a pressure wave of of a trial waveform with a long dwell time is proposed. By using
the ink inside the inkjet head. In the paper by Bogy and Talke, the test waveform, meniscus motion from either the rising or
meniscus-protrusion behavior was discussed in relation to pres- the falling region of the waveform can be analyzed separately.
From the period of the measured motion, the optimal value
for the dwell time of the waveform can be determined. Dwell
Manuscript received February 2, 2009; revised April 30, 2009. First time is known to be one of the critical parameters in designing
published July 31, 2009; current version published September 30, 2009.
Subject Editor Y. Zohar. a waveform [6]–[8]. The proposed waveform design scheme
The author is with the Department of Mechanical Engineering, Soonchun- will be extended to a waveform for high-frequency jetting; this
hyang University, Asan 336-745, Korea (e-mail: kskwon@sch.ac.kr). requires two optimal values for dwell time since there are two
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org. dwell components in the waveform. Finally, the jetting behavior
Digital Object Identifier 10.1109/JMEMS.2009.2026465 of a negative waveform will be discussed.
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
KWON: WAVEFORM DESIGN METHODS FOR PIEZO INKJET DISPENSERS BASED ON MEASURED MENISCUS MOTION 1119
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
1120 JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 18, NO. 5, OCTOBER 2009
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
KWON: WAVEFORM DESIGN METHODS FOR PIEZO INKJET DISPENSERS BASED ON MEASURED MENISCUS MOTION 1121
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
1122 JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 18, NO. 5, OCTOBER 2009
Fig. 8. Meniscus motion from the negative trial waveform with long dwell
time (v1 = −17 V and t1 = 100 µs). (a) Meniscus behavior from the negative Fig. 9. Meniscus behavior from the optimal waveform.
trial waveform. (b) Measured meniscus motion.
the meniscus motion of the jetting material is small, a higher
use of a negative trial waveform can have advantages since waveform voltage may be needed for jetting.
the first measured meniscus motion is likely to be the result Note that a dwell time that is equal to the period of menis-
of the pressure wave generated from the first falling section cus motion can cancel the pressure waves since the phase
of the waveform only. In the previous positive trial voltage, of the pressure waves from the rising and falling sections of
shown in Fig. 7, the protruded-meniscus motion due to the a waveform is 180◦ . As shown in Fig. 5, the droplet speed
final falling section of the waveform may be subject to the first was reduced significantly when the dwell time approached 40,
rising section, even though the meniscus oscillation from the 80, 120 µs, etc., which is repeated due to the periodicity in
first rising section is damped. meniscus behavior.
Once the period of meniscus motion from either the rising or
falling section of the waveform is measured, the optimal dwell
C. Waveform Design for High-Speed Jetting
time can be easily determined. Assuming that the pressure
waves (or meniscus oscillation) from the rising and falling For reliable jetting, a subsequent droplet should not be
times have the same period with a phase difference of 180◦ , ejected until the meniscus vibration from the previous drop has
the positive meniscus can be maximized by setting the dwell sufficiently decayed. This damping takes time and limits the
time to a one-half period of the measured meniscus motion, as maximum jetting frequency. High-frequency jetting is required
shown in Fig. 9. Therefore, the optimal value for dwell time to increase the speed of printing. To obtain higher frequency
was determined to be 20 µs (t1 = T /2), which is in good jetting, a proper waveform for the dispenser needs to be de-
agreement with the optimal dwell time from the speed and signed such that residual meniscus oscillation can be effectively
dwell-time relationship shown in Fig. 5. The jetting behavior suppressed after droplet jetting [14].
can then be predicted without actual jetting by measuring the Compared to a basic waveform, more parameters, including
meniscus motion. For example, Fig. 6 shows the meniscus two optimal dwell times, are required to determine a waveform,
motion from the waveform with optimal dwell time t1 = 20 µs. such as that shown in Fig. 10 for high-frequency jetting [6], [9].
By observing the meniscus motion, we see that jetting will The first dwell time t2 is related to jetting, as discussed in the
likely occur at 60 µs after the jetting signal is applied. Later, the previous section. In addition, it was found that the optimal value
second meniscus peak at 100 µs is likely to become a satellite for maximizing jetting pressure is one-half of the measured
drop if the magnitude of the waveform becomes large. meniscus period from the trial waveform; i.e., t2 = T /2. The
As a final step in waveform design, the magnitude of the second dwell time t3 is related to the suppression of residual
waveform was determined such that the target droplet jetting meniscus oscillation after jetting.
speed can be satisfied. The magnitude of meniscus motion is The effect of the first dwell time t2 can be understood from
related to jettability of the jetting materials. If the magnitude of the conventional approach using droplet jetting speed, as shown
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
KWON: WAVEFORM DESIGN METHODS FOR PIEZO INKJET DISPENSERS BASED ON MEASURED MENISCUS MOTION 1123
Fig. 12. Meniscus behavior from the optimal waveform for high-frequency
jetting.
second dwell times (t3 = 20 and 80 µs). Note that the second
dwell section of the waveform is unlikely to affect jetting,
as seen in Fig. 11.
As a final step in designing the waveform, appropriate volt-
ages in the waveforms v2 and v3 will be determined based on
the measured jetting speed and residual meniscus oscillation.
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
1124 JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 18, NO. 5, OCTOBER 2009
ACKNOWLEDGMENT
The author would like to thank J.-H. Myung for the assis-
tance with experiments.
R EFERENCES
[1] D. Albertalli, “Gen 7 FPD inkjet equipment—Development status,” SID
Symp. Dig. Tech. Pap., vol. 36, no. 1, pp. 1200–1203, May 2005.
[2] D. Redinger, S. Molesa, S. Yin, R. Farschi, and V. Subramanian, “An ink-
jet-deposited passive component process for RFID,” IEEE Trans. Electron
Devices, vol. 51, no. 12, pp. 1978–1983, Dec. 2004.
[3] S. B. Fuller, E. J. Wilhelm, and J. M. Jacobson, “Ink-jet printed nano-
particle microelectromechanical systems,” J. Microelectromech. Syst.,
vol. 11, no. 1, pp. 54–60, Feb. 2002.
[4] J. E. Fromm, “Numerical calculation of the fluid dynamics of drop-
on-demand jets,” IBM J. Res. Develop., vol. 28, no. 3, pp. 322–333,
May 1984.
[5] N. Reis, C. Ainsley, and B. Derby, “Ink-jet delivery of particle suspensions
Fig. 16. Measured meniscus motion from the negative waveform by piezoelectric droplet ejectors,” J. Appl. Phys., vol. 97, no. 9, p. 094 903,
(v4 = 19 V). Apr. 2005.
[6] “Drive waveform effects on ink-jet device performance,” MicroFab
Technologies, Inc., Plano, TX, MicroFab Technote 99-03, 1999.
the meniscus into the nozzle excessively in some cases; e.g., [7] D. B. Bogy and F. E. Talke, “Experimental and theoretical study of wave
t4 = T /2, 3T /2, etc. propagation phenomena in drop-on-demand ink jet devices,” IBM J. Res.
Develop., vol. 28, no. 3, pp. 314–321, May 1984.
[8] J. B. Szczech, C. M. Megaridis, and D. R. Gamota, “Fine-line conductor
manufacturing using drop-on demand PZT printing technology,” IEEE
IV. D ISCUSSION AND C ONCLUSION Trans. Electron. Packag. Manuf., vol. 25, no. 1, pp. 26–33, Jan. 2002.
[9] J. Chung, S. Ko, C. P. Grigoropoulos, N. R. Bieri, C. Dockendorf, and
New measurement techniques have been presented to mea- D. Poulikakos, “Damage-free low temperature pulsed laser printing of
sure inkjet meniscus motion using a CCD camera system gold nanoinks on polymers,” Trans. ASME, J. Heat Transf., vol. 127, no. 7,
with strobe LEDs. Jetting behavior can be predicted from the pp. 724–732, Jul. 2005.
[10] B. Beulen, J. D. Jong, H. Reinten, M. V. D. Berg, H. Wijshoff, and
meniscus motion. To analyze the waveform effect on meniscus R. V. Dongen, “Flows on the nozzle plate of an inkjet printhead,” Exp.
motion, the use of a trial waveform with a long dwell time was Fluids, vol. 42, no. 2, pp. 217–224, Feb. 2007.
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.
KWON: WAVEFORM DESIGN METHODS FOR PIEZO INKJET DISPENSERS BASED ON MEASURED MENISCUS MOTION 1125
[11] C. D. Meinhart and H. Zhang, “The flow structure inside a microfabricated Kye-Si Kwon received the B.S. degree in mechanical
inkjet printhead,” J. Microelectromech. Syst., vol. 9, no. 1, pp. 67–75, engineering from Yonsei University, Seoul, Korea, in
Mar. 2000. 1992, and the M.S. and Ph.D. degrees in mechanical
[12] K. S. Kwon, “Speed measurement of ink droplet by using edge detection engineering from Korea Advanced Institute of Sci-
techniques,” Measurement, vol. 42, no. 1, pp. 44–50, Jan. 2009. ence and Technology, Daejeon, Korea, in 1994 and
[13] K. S. Kwon, Meniscus Motion Measurement2009. [Online]. Available: 1999, respectively.
http://www.youtube.com/watch?v=U48a7bRtJ9I He has been an Assistant Professor in the Depart-
[14] K. S. Kwon, “A waveform design method for high-speed inkjet printing ment of Mechanical Engineering, Soonchunhyang
based on self-sensing measurement,” Sens. Actuators A, Phys., vol. 140, University, Asan, Korea, since 2006. Before joining
no. 1, pp. 75–83, Oct. 2007. Soonchunhyang University, he was a member of the
[15] M. B. G. Wassink, M. J. M Bosch, O. H. Bosgra, and S. Koekebakker, Research Staff at Samsung Advanced Institute of
“Enabling higher jet frequencies for an inkjet printhead using iterative Technology, Yongin, Korea. His current work includes ink jetting control.
learning control,” in Proc. IEEE Conf. Control Appl., Toronto, ON,
Canada, Aug. 28–31, 2005, pp. 791–796.
Authorized licensed use limited to: Soonchunhyang University. Downloaded on October 4, 2009 at 06:25 from IEEE Xplore. Restrictions apply.